Discussion of Session 2: Toxicological and Environmental Aspects Related to the Use of Plant Oils as Fuels

Transcription

1 Discussion of Session 2: Toxicological and Environmental Aspects Related to the Use of Plant Oils as Fuels Presentation by J Mauderly: Health Issues Concerning Inhalation of Petroleum Diesel and Biodiesel Exhaust. F Adlkofer:, Dr Mauderly, you said that you don't think health effects might be the limiting effects in introducing Biodiesel. I think that is absolutely correct. The question is: what about a lower risk resulting from Biodiesel as compared with fossil diesel? Would this not be a reason for introducing it, to improve the economic situation? J Mauderly: Logically, yes. But politics are not always logical. If the question is posed differently and one asks: if all diesel engines in the world were running on Biodiesel rather than petroleum diesel, would this be a better world? It's difficult to say, but probably, yes. The next question is: how much are you willing to pay for this better world? And my answer is: I'm not sure, I'd have to ask my colleagues, and how much money are we really talking about? The answer really depends on how concerned one is about the risk of petroleum diesel exhaust This is more political than scientific, so I cannot give a very good answer. Presentation by J Krahl: Biodiesel Exhausts Emissions and Determination of their Environmental and Health Effects. G Oberdiirster: What kind of cell is the L929 cell that you used for toxicity testing? J Krahl: A mouse fibroblast cell line. G Oberdiirster: How do the results compare with other cell types that are used for toxicity testing? J Bunger: We did the same test with L 132 cells, an established embryonic lung cell line, with the same result J Mauderly: With respect to mutagenicity, it was not clear, whether you meant specific mutagenicity, i.e. mutagenicity per microgram of extract, or total mutagenicity, i.e. mutagenicity per hour of use of engine or per mile travelled or something like that? What were the units?

2 182 J Krahl: The unit was the volume of exhaust. The exhaust volume is nearly the same whether an engine is run on diesel or Biodiesel, although the mass can be completely different Exhaust gas from RME, as in the example of the French tractor, can have double the mass of fossil diesel exhaust. We always took a volume and extracted a volume. J Mauderly: So this would be a comparison of mutagenic activity per unit of equivalent engine use. G Oberdiirster: The particle size distribution measured by an impactor showed a shift towards larger particles. Did you make any electron micrographs to see if these particles consisted of agglomerates. If so, what kind of agglomerates were they, and what were the smallest particles? J Krahl: This is being done currently in Magdeburg. The electron micrographs show the agglomeration of Biodiesel particulate-this is the first result-to be higher than with fossil diesel. Thus we find the same tendency with microscopy as well. G Oberdiirster: So the primary particles may be of the same size between the two exhausts, but Biodiesel exhaust agglomerates more easily. J Krahl: Yes. F Adlkofer: May I return to the mutagenicity related to the volume of exhaust? If you need more Biodiesel to go a certain distance because it is less energetic than fossil diesel, Biodiesel might therefore generate a higher volume, and the total mutagenicity resulting from Biodiesel might be even higher than with fossil diesel. J Krahl: The energetic fuel consumption is the same for Biodiesel and diesel. The volumetric consumption is 8 % higher and the gravimetric consumption is 13 % higher, whilst the differences in the mutagenicity tests between diesel and Biodiesel are larger than 13 %. J Schell: Do you have any recommendations for the plant breeders on the basis of your data? J Krahl: The transparency showed the technical procedure-measurement, assessment, technical consequences. At the moment we are between the first and second steps and we are not sure what to recommend Presentation by P Eckl: The Mutagenic Potential of Diesel and Biodiesel Exhausts. G Oberdiirster: My question concerns the dosing of the hepatocytes. You expressed the dose with respect to exhaust volume. What is the mass dose given to

3 183 your cell system? For 50 I exhaust volume, for example, how much of that is particulate and organic mass? P EckI: The particles have to be extracted when we perform the assay. We get a total sample volume of some 10 ml, corresponding to about 1-2 m 3 Recalculating for 1 1, gives a maximum of about 50 ml sample volume or so. For the dose-response curves, up to a maximum of 5 % was used in the test system, i.e. 5 % of the volume, making it picogram per 500,000 cells. J Connemann: Could you describe the ester that was tested? If, for instance, it contained considerable glycerol, the exhaust gases will have very high content of acrolein. It's very difficult to compare the different results of toxicity testing. It is important to know what sort ofbiodiesel being tested. P EckI: That is why we compared directly the Ames Test and the hepatocyte assay with the same sample, so as to have the same concentration of substances present. J Connemann: The results are very different with a high content of methanol or a high content of glycerol. The limit for total glycerol content here is For the USA Mr D. Reece showed me values from 1.80 to The best one was 3 times above our limit. J McDonald: This is correct, although I have seen much better quality esters. My data are from a very pure double vacuum-distilled ester. However, I have also participated in a field trial in an underground mine. I had to return 2,000 gallons of fuel to the supplier because of contamination with glycerol, glycerides, and water. Ethanol and methanol, moreover, are aldehyde precursors, and in diesel combustion they could combust partially to form formaldehyde and acetaldehyde. H PrankI: In reply to Mr Connemann. The fuel was produced at our institute. The fuel specifications met DIN standard (except neutral number and water content). Presentation by J McDonald: Biodiesel: Effects on Exhaust Constituents. J Mauderly: If I understood correctly, you said that in your studies the greater soluble organic fraction of soot was attributed almost exclusively to a very narrow set of unburned fuel constituents, linolenic, linoleic, oleic acid esters. J McDonald: Yes. Of the mass of the organic material collected on the filter, 55 % was contributed by those compounds. J Mauderly: Given other data. do you think that is characteristic? J McDonald: Yes, and it may have a lot to do with the variability in particulate matter emissions from engine to engine. Under some conditions organic material

4 184 might exit in the vapour phase, under other conditions in the particulate phase, depending on the dilution tunnel and the dilution ratio used. It also depends on the fatty acid profile of the Biodiesel fuel used. J Mauderly: Would you then predict that if the plant and fuel processors were clever enough to arrange the hydrocarbons on either side of that peak, the organic fraction would be markedly reduced? J McDonald: Yes, if it were arranged on the lower carbon number side of that peak. A careful examination of the toxicity of methyl esters and the particle phase is warranted. If the material is relatively benign, what difference does it make? It makes a difference from the standpoint of regulated emissions, because regulated emissions are defined by a temperature, 52 e, dilution with air, and deposition on a specific type of filter. Dilution ratios typically used in laboratories are between about 10: 1 and 15: 1. In view of this and a comparison with dilution in ambient air, there would appear to be significant differences between a particle in a laboratory dilution tunnel and a particle in ambient air. Our tests in underground mines show just this. Just because organic material is present in the particle phase in the dilution tunnel doesn't mean that it will ever enter the lungs as particulate matter, because it may just evaporate once it's diluted more with air. And even if it does, it need not be harmful. F Adlkofer: If compared with the data of Dr Eckl and Dr Krahl, your data would appear to favour the future use ofbiodiesel. What is the reason for that? J McDonald: I defer to Dr Mauderly. It is not worse with respect to indicators of potential health effects. F Adlkofer: Did you use Biodiesel from a different source? J McDonald: Soybean oil. The differences include, firstly, the indirect injection engine. In our experience, indirect-injection engines are much more tolerant of fuels with different physical characteristics. Those two particular indirect injection engines can operate on high percentage soy methylesters, but recent testing of a very modem direct-injection Navastar engine demonstrated that it could not run with much more than 50 % soy methyl ester without reliability problems. Regarding the emissions, so many factors influence the measurement of emissions results. I would need to see the error bars, to see the uncertainty level. Factors such as the fuel, compounds in the fuel, error or statistical uncertainty can influence the outcome. There have been frequent reports on carbon monoxide reductions, but carbon monoxide is not a problem with diesel engines. There are no efforts to reduce carbon monoxide emissions in general with diesel engines, because these are very low. The question arises: for those emissions that currently are a concern with diesel engines, such as P AH, are the concentrations above the minimum detection limit, and if so, are they still really low? G Oberdorster: How old were the engines you used, how long had they run?

5 185 J McDonald: The MWM engine was brand new, but it's an older design. Many of the engines used were older. G Oberdorster: Do emissions change with engine age? J McDonald: The same engine design is produced today. The engine was about 5 years old, but had run for less than 100 h. The Caterpillar engine was recently rebuilt It had run for about 1500 h, and was rebuilt to brand-new specifications. G Oberdorster: Could you compare emissions of an aged engine with those of a brand-new one? J McDonald: An aged engine has increased organic emissions, increased oil consumption, increased particulate matter emissions, and the oxides of nitrogen emissions decrease. There are differences: both those engines were well broken in. Do you mean a brand-new engine off the factory floor? Or a newer engine design? The particulate matter emission for the Caterpillar engine was around g/ kwh, which is very high. Modem engines achieve typically 0.1 g/kwh, and even this can be reduced. The Series 60 and Series 50 Detroit Diesel engines for instance, which have been tested extensively by South West Research Institute, have shown very similar reductions in total particulate matter when used with a diesel oxidation catalyst to these older style engines, although I would expect the particle size distribution from those more modem engines to be somewhat different. They may have more nuclei mode particles. Presentation by W Rodinger: Toxicology and Ecotoxicology ofbiodiesel Fuel. G Oberdorster: What was the "inhalation" endpoint in your data? W Rodinger: The inhalation data I found in the literature. G Oberdorster: On Biodiesel fuel? W Rodinger: On fuel and on diesel exhaust. G Oberdorster: So that's combined, exhaust plus fuel. J Mauderly: Dr Rodinger, did you see any biological endpoint in any system that gave you cause for concern? In all the many measurements there was not one in which the Biodiesel was worse than petroleum diesel. W Rodinger: I found only one with macrophytes. The Ashton research had found Biodiesel to be more toxic to macrophytes than mineral-based diesel. J Mauderly: What might the reason for this be?

6 186 W Rodinger: The litemture suggests that there is a physical or mechanical effect which influences toxicity more than the chemical effect. G Oberdiirster: In the final part of this discussion, the major concern is questions that can be addressed to the plant bioscientists with regard to improving the product Biodiesel and its exhaust, in terms of its toxicological and ecotoxicological impact? From what has been said, Biodiesel looks very good. Particulate emissions from Biodiesel fuel were basically less toxic than.petroleum diesel and there was also a significant decrease in P AH. The only increases for Biodiesel reported in different studies have been NOx emissions and perhaps aldehydes. In the first talk, Dr. Maudedy emphasised three diesel exhaust components that should be evaluated, namely particulates, PAH and irritant compounds. We will discuss the toxicology first, and then move to the ecotoxicology and try to formulate specific questions for the plant scientists. E Garshick: It might be relevant to review the potential health effects in humans, some of which were addressed by Dr Maudedy. Odour and irritant effects were mentioned. The relevance of these will depend on the presence of irritants such as acrolein, and given the variable fuel glycerol content, the presence of N02 Respiratory symptoms such as cough, phlegm and wheeze may occur in exposed individuals. Under certain conditions the fuel and engine opemting conditions may also be important. Studies of acute exposure to fossil fuel diesel exhaust have shown pulmonary function changes in exposed individuals, although not consistently so. This might be relevant for individuals exposed by opemting vehicles powered by Biodiesel fuels. For fossil fuel diesel, there is little or no information on the effects of chronic exposure on pulmonary function. There are no long-term longitudinal studies. We should consider this if a new industry is to be developed, and start to make such measurements in individuals who may be exposed subsequently on a long-term basis. It has been suggested that bladder cancer can develop in workers exposed to P AH such as occurs in fossil fuel diesel. However, the epidemiologic litemture linking fossil fuel diesel exhaust to the occurence of bladder cancer is not conclusive. Motor vehicle opemtors do show an increased incidence, but it's unclear whether this can be attributed to diesel exhaust exposure since exposure in these studies has not been well chamcterized. Epidemiologic studies have also suggested that there is a small increase in lung cancer risk in workers occupationally exposed to diesel exhaust over many years. Speaking as an epidemiologist and physician, a lesson relevant to the Biodiesel industry is to characterize emissions thoroughly since emissions can be influenced by the specific fuel being used. It is important to determine an appropriate emission marker to be used for environmental sampling. In studies on fossil fuel diesel exhaust one factor limiting the interpretation of epidemiologic studies is the lack of a marlcer specific for diesel emissions. Recent work has indicated that elemental carbon is a marker of emissions for fossil fuel diesel exhaust. This might be an appropriate marker for Biodiesel based exhaust emissions, or biomonitoring using new technology may be possible. In the Biodiesel industry relatively few individuals are exposed as yet. It would be useful to start prospective studies in workers with the extent of exposure well-chamcterised in

7 187 order to study the potential health effects. Finally, surveys assessing respiratory symptoms and pulmonary function should be conducted among the individuals whose exposure is well characterised. These surveys must be carried out over many years, particularly in view of the long latency of cancer, which is usually more than 20 years for lung cancer. A potential nuuor confounder in any study of respiratory function in an occupational cohort would be smoking habits, and I would urge that smoking habits be considered prospectively. G Oberdiirster: Such studies are certainly important, but at the moment the population in which such studies could be performed is not large enough. As important as such studies are, until there is a large enough cohort we should to focus on chamcterising the toxicity and carcinogenicity which might be involved with Biodiesel. Even if, for instance, particulate emissions are lower from Biodiesel, the effects need not necessarily be less. However, our information to date suggests that Biodiesel exhaust has a lower mutagenic, carcinogenic and potentially also irritant potency. However, we cannot be certain until studies have been performed in a mammalian species. J Mauderly: Overall nothing has suggested strongly that health concerns will limit the use of Biodiesel. This is a very broad statement, but probably a good starting point I would encourage the plant scientist or the fuel processor or the engine manufacturer as follows: if a very narrow family of hydrocarbons contribute the majority of the organic mass in exhaust, then something may be gained by trying to reduce the content of that particular family of compounds in the fuel. Why? Even if this organic material is not strongly mutagenic, if it contributes up to 50 % of the total organic, which may be 20 % of the particle mass, then reducing it would reduce particle mass concentrations even further. With respect to work that ought to be done on health effects, there is a difficulty in that information has already been obtained from many different kinds of fuels and engines. For a health or laboratory scientist beginning a more definitive evaluation it would be highly desirable if some typical condition were to emerge as representative. Will there be a typical fuel or a most typical engine? Until the industry progresses to the point where there some uniformity, biological results will vary tremendously depending upon the conditions. Assuming that "typical" exhaust generation conditions could be determined, what then? One gap in our current understanding that would be very straightforward to resolve would be a simple animal inhalation study to compare the irritant response. Although extrapolation from animal studies to humans is not without problem, comparative studies of a certain dilution of petroleum diesel exhaust or a Biodiesel exhaust could still be performed in rodents to see whether the response to the total diluted exhaust-the vapour phase and the particle phase -is really noticeably different. Such a study would be neither time consuming nor expensive nor hard to do. G Oberdiirster: Something should probably be done in a mammalian species to compare the two exhausts. Separating the gas phase may not necessarily be immediately necessary, but comparative data with respect to toxicity, inhalation

8 188 toxicity are needed to establish whether Biodiesel may be of less concern than petroleum diesel, or how they differ. The point about a standardised fuel is a vel)' good. A DIN nonn for Biodiesel has been discussed: will it be good enough to have a standardised fuel? Can this be a basis for testing purposes? J McDonald: That is my suggestion. Diesel fuel, petroleum diesel fuel, is by no means homogenous. One thing that has been done in testing petroleum diesel fuels and fuel additives has been to standardise a test fuel with tighter specifications. This is done for certification purposes allover the world. Both the USA and the EU have standardised test fuels with specific properties that are representative of a typical fuel, but allow much closer comparison of different exhaust properties and much more repeatable results. I would thus recommend a test specification for a test fuel that is typical and is more stringently controlled than the biological material obtainable, which can v8l)' considerably from season to season and from processor to processor. H Klingenberg: To return to the health effects on humans and the threshold Dr Maudedy mentioned. The air quality values in a street with dense traffic and quite a high number of diesel cars, show some 10 ~g/ m 3. If the fleet were to have 20 % Biodiesel cars, an optimistic value, the concentration would fall to about 9 Jlg/m 3 With the impact being far below the effective threshold, changing the fuel will have practically no effect on humans. G Oberdorster: That is certainly correct for the overall exposure situation, but if the Biodiesel is used specifically in certain areas where people are exposed only to Biodiesel, this would be different. H Klingenberg: That is a vel)' limited application. For instance in mines, where diesel engines have to be used, the particulate concentration can be reduced, but it must be realised that there are also other effective particulates that are present in far, far higher concentrations than diesel particulates. J Mauderly: If I may respond briefly. Professor Klingenberg referred to my use of the tenn threshold. I want to make clear how I was using that teno, namely in relation to the animal studies, in particular the rat for which there seemed to be a threshold of some sort for both the lung tumour response and the non-cancer lung irritation response. I do not know if threshold exists for human responses, and when discussing environmental exposures at the concentrations mentioned discussing, this is vel)' relevant In this situation it is not simply exposure to diesel exhaust only, but exposure to ambient particulate material, which is a far broader issue. Within ambient particulate material, diesel, Biodiesel would be just one of many constituents. The threshold issues would be quite different at those concentrations. H Klingenberg: I am aware of the threshold, and I was going to ask about the current state of the discussion. I have followed the question of thresholds for many years and indeed it is not known if there is one for humans. Referring to the

9 189 particulates again, a concentration change of only 1 Jig! m 3 is 10 0/0, and considering the other particles present at concentrations far higher than that from the diesel cars. my argument still holds true. G Oberdorster: There may be another argument of importance: The US-EPA is trying to introduce a so-called 'Fine Particle Standard'. Within the fine particle mode there is the nucleation mode of ultrafine particles, which may be very importaril The mass of this mode is not so relevant because it is very, very low, so even 1 Jig represents a large nwnber of particles. J Schell: Molecular breeders can achieve something only if the task is clearly defined. It has been suggested that most of the organic particulate matter originates from only a few organic products. Can these be identified? J McDonald: In my studies they have been identified. They were primarily methyl esters and oleic and linoleic acids. J Schell: This is where the molecular biologists and breeders in geneml could do something about it. J McDonald: I am aware of this. However, when discussing points of no return and No,. emissions the ignition delay effect must be considered. Evidence from research in the USA shows that going to exceptionally short ignition delays with exceptionally high cetane nwnber fuels may actually increase oxides of nitrogen emissions. Since there has been very little combustion work done with Biodiesel it is not clear if this is also the case with Biodiesel fuel. It is conceivable that the increase in cetane number above 60 in Biodiesel fuel might be contributing to the detrimental effect on oxides of nitrogen emissions. Thus it's not necessarily a good idea to go to even higher cetane nwnber fuels. G Oberdorster: You mentioned specifically a change in condensation and absorption characteristics with respect to the particulate output. J McDonald: We did some initial modelling of the condensation and absorption characteristics. G Oberdorster: What are the implications for specific fuel constituents? J McDonald: Carbon number. Again carbon number. Lower molecular weight fatty acids would be less likely to condense and absorb and other things are affected also. The degree of saturation and branching of the molecules also affects condensation and absorption characteristics. G Oberdorster: So lowering the carbon chain number is a desirable trait? J McDonald: Yes, but organic material is not the entire problem. Methyl esters appear to be fairly benign, and although they may condense and absorb in EPA or

10 190 EU-style laboratory dilution tunnels, there is little or no evidence for their condensation in ambient air. One might ask: why go to all the effort of modifying the fuel if the organic material in the ambient does not condense or absorb to form particulate matter at the high dilution ratios present in the atmosphere. G Oberdiirster: So you see no need to change anything at the moment? J McDonald: No. If the problem is meeting a particular particulate matter standard, this has already been achieved compared with petroleum diesel fuel. G Oberdiirster: What do you call a standard? J McDonald: In the EU it is 0.15 glkwh. If a particular engine already achieves that with Biodiesel, or is no worse than with fossil diesel, it doesn't need to be reduced. G Oberdiirster: This is a general standard for emission, but it may not necessarily mean that it wouldn't be good to reduce it further if possible. J McDonald: It's possible, but if it's not condensing into particulate matter in the ambient air, is it a hazard to health? W Rodinger: Particulate matter from fossil diesel is always being compared with that from Biodiesel. In principle it is incorrect to compare gravimetric values: we conclude that the value is higher, and that the effect is worse. Investigations have shown that the composition of the particulate is completely different and the effect of the particulate may be different. Comparing the fuels on the basis of gravimetric value is like comparing oranges with apples. There has also been a proposal for shorter fatty acid chains in the fuel. If this were done, I suspect that aldehyde emissions would increase, because the aldehyde emissions may be connected with the ester group in the fuel. Shorter chains mean more molecules for the same heating value, and so the emission of toxic aldehydes may increase. G Oberdiirster: Your comment on particles is correct; all particles are not necessarily the same. Although masses are not always the same, what else can be used? Toxicology employs other dosimetric indices such as surface area This is very important for agglomerates of very fine particles for which surface area could be very large; this has an impact on the effects. However, without animal studies it is hard to predict what the effects of Biodiesel will be compared with petroleum diesel. Toxicological studies in that regard are necessruy. It may be predicted that reduced particle emissions by mass are beneficial, but this is not certain. P Walzer: Many of the results relating to particulates and NOx have been gained from today's not very advanced diesel engines. Future emission legislation will require that diesel particulate emission be almost as low as that of gasoline engines and that NOx be reduced considerably. To achieve this, the same improvement step that was made in emissions from 1975 till today must be

11 191 repeated by the year , especially in the USA. Diesel engine technology is developing in the direction of high-pressure, multi-injection and pre-injection, and even to control of combustion process. NOx will be reduced by very high degree of exhaust gas recirculation. All this will change the emission constituents. So in a certain respect, we have been discussing a status which is already past. For instance, it was suggested that the diesel should perhaps have a pre-chamber etc. This is, in other words, attempting to choose an engine according to the fuel. Engine development is however proceeding in other directions. It would thus be preferable to take such a very advanced diesel engine and run it on Biodiesel and so determine the direction of changes in emissions. Perhaps Biodiesel can even help under these conditions to meet future emission legislation requirements. G Oberdorster: This problem occurs constantly. Whenever research is undertaken, the engine is already outdated. P Walzer: Outdated is the wrong term, the direction of engine development is changing, and ifbiodiesel is to be a fuel for all engines, the combustion processes of these advanced diesel engines must be studied. This is another area where information is inadequate. F Adlkofer: Dr Walzer is implying that Biodiesel should be treated in the same way as fossil diesel, because the problems might be solved by engine technology. The outcome for the environment and for humans exposed to this diesel would be exactly the same if the problems are solved by changing the engine. The second point is NO,/N02. Plant biologists can't do much to reduce NOx and N02. It comes from the air and is created when Biodiesel is burned. G Oberdorster: Your point was that when an improved Biodiesel engine is available, the exhaust constituents and their effects should again be characterised. F Adlkofer: The consequence is that there is no advice to be given to the plant biologists at present. There are possibilities in changing the engine, and that is valid also for the normal diesel. It means that both could be treated similarly, including the N02INOx problem. That would be solved by engine technology in the same way as with fossil diesel. P Walzer: I cannot predict how Biodiesel will perform in such a high-pressure injection, high exhaust-recirculation environment. Hopefully better, but we should base our first research on these results. The first and urgent step is to investigate Biodiesel in such a modern engine. J McDonald: To some extent such investigations have already been made in the USA, at least using soybean-based diesel fuel, mostly in blends with petroleum diesel fuel. To my knowledge there have been no studies on high-pressure injection systems and highly-boosted direct-injection engines with either pure or a high percentage of rape methyl ester.

12 192 G Oberd6rster: There are data for soybean methyl ester? J McDonald: For blends, and limited data for pure soy methyl ester on Detroit Diesel Series 50 and Series 60 engines, and I think also on the Caterpillar 3406E. These engines are pretty much state-of-the-art as far as heavy-duty engines are concerned. G Oberdiirster: What do the results suggest? J McDonald: With respect to the carbon fraction of particulate matter the results are fairly similar. Carbon particulate matter is reduced more or less in parallel with the oxygen content of the fuel, so blending Biodiesel with petroleum diesel reduces the carbon fraction and tends to increase the organic fraction. G Riibbelen: Dr McDonald, you stressed a test fuel. In what way is a test fuel different from a nonn? The test fuel also has to be defined, one simply needs the details of the qualities present. Whether this is a test fuel or a written norm, is this different? If not, a standardised blend may be more realistic and immediately useful than a nonn. But otherwise a norm is as good as a test fuel, isn't it? A Schifer: No. Consider density, for diesel purposes, as an example. EN 590, the existing diesel standard in the EU, demands densities of kglml. In comparison, the reference fuel that must be used to certify our engines, for example, allows just , i.e. just one-sixth of the permissible range of densities found on the market. This enables engine-typical effects to be distinguished from fuel-typical effects. All properties of a reference fuel have to be defined far more stringently than the normal standard for market use. J McDonald: Consider the current range of fuel sulphur contents. Testing fuel additives in the U.S. requires a sulphur content of 0.05±O.00250/0; engine certification testing requires a sulphur content of %. There is a range of aromatic contents that has to be adhered to. For diesel fuels, for certification fuels, there is a whole range of much tighter specifications than are required for overthe-road vehicles. A good example would be the use of on-highway diesel fuel for certification testing in the USA. Everyone would use zero-sulphur fuel-a legal on-highway fuel-that would result in much lower particulate emissions, and would show the engine to be somewhat greener than it really is. P Eckl: From a toxicological point of view, Biodiesel seems better than diesel, except for few discrepant results like mine. But uncertainties remain, in particular because there are no in vivo data and hence no direct comparison can be made with fossil diesel data. Thus there is a demand for more research in this area Furthennore, from a toxicological point of view, if the engine designers or fuel designers agree that longer-chain, or shorter- or branched-chain fuels are preferable, the toxicologist can evaluate afterwards the advantages or otherwise. The fuel must first be designed, then it can be evaluated and consequences drawn.

13 193 G Oberdorster: That is, both the fuel and the improved engine are necessary before toxicological testing can be started. F Adlkofer: This direction presents a big problem. It is really not known which substances in Biodiesel exhaust are the most dangerous, nor is it known which substances have to be removed from the exhaust to make it safe for the environment and exposed animals and humans. G Oberdorster: General suggestions are possible. It's always beneficial to reduce particles as far as possible, similarly P AH and irritant compounds. However the extent to which this is necessary is unknown. F Adlkofer: These principles are without doubt correct. Although a little is known about these substances, their role in diesel exhaust toxicology is unknown. G Oberdorster: Because there are no in vivo data on Biodiesel toxicology. F Adlkofer: Ifwe look at normal diesel? G Oberdorster: The rat studies suggest that particles are in all likelihood the culprits for induction of the tumours in the rat P AH is not needed to induce tumours in rats with high exposure concentrations. F Adlkofer: This is exactly the problem. Whether P AH plays a role at all in these animal studies in which lung tumours were caused by diesel exhaust is not known, the particulate matter has been identified as the culprit Extrapolating to the present situation, should we recommend that the plant biologists do everything possible to reduce particles? We don't even know the difference between particles from fossil diesel exhaust and those from Biodiesel exhaust G Oberdorster: This is correct. The present dilemma is that we don't know what should be recommend or even what questions should be asked. J Mauderly: To summarise, there are limited data on the biological activity of Biodiesel exhaust Importantly, the biologists are not waving a flag and saying there is something terribly dangerous here. This being so, only one clear recommendation can be made to the plant biologists: make plants that make more oil, and make it economically feasible. When a viable industry has been established, and fuels and engines become more standard, toxicology can be revisited. G Oberdorster: In short, the plant biologists must increase the crop yield.